Slightly Advanced Statistics

With about a quarter of the season already played (!!), and the Sharks sitting atop the Pacific, 3rd in the West, and 7th overall*, I thought I'd dust off the old Microsoft Excel, dig up some W/L/OTL stats, and see how the Sharks have been doing this season, as well as how they've played compared to some other teams.

First up, let's take a look at how Pacific Division teams have collected points through 19 games...

 ... not terribly enlightening, unless you're really interested in knowing how many points the Coyotes had after playing 11 games. This chart overlaps a little bit too much, and due to the nature of the data, this chart doesn't tell us a whole lot. I guess it's kinda useful for looking at long winning/losing streaks, a la the Stars from games 2 to 7 or the Stars from games 14 to 19. But even long streaks are just represented as a line with positive slope or a flat line (for winning streaks and losing streaks, respectively). So let's move on to some more meaningful data... or rather, data manipulated in a more meaningful way...

It's been a while since I've futzed with Excel, so inputing just the California teams took long enough, so those are the only ones represented on this chart. Here I did a simple points divided by games played calculation (thank god for =column/column) to basically see how efficient these teams were at collecting points. Pts/GP is basically Points% multiplied by two. Personally I prefer it because 1)it's how I happened to do these calculations and 2)it offers an easier way of looking at wins, losses, OTL just from a number. A line up at 2 indicates only wins, a line at 0 would be no wins, and a line at 1 either means an overtime loss or a win and loss. Ideally you'd want your team up at the 2 mark, but realistically that's not going to happen in real life (...my team in NHL12, that's another story), so a good mark to shoot for would be about 1.241 (the average Pts/GP required last season to make the playoffs, unless you're Dallas or Calgary, who both had more points than the Rangers but missed the playoffs). As a benchmark, last season the Sharks Pts/GP was 1.28 and the Pacific Division averaged 1.209. Currently, at 19 GP, the Sharks Pts/GP is 1.421, which is fantastic, but also probably something that's going to drop over the course of the season (last season, the highest Pts/GP was Vancouver's; 1.427). As a comparison, the Kings (with 22 GP) are sitting at 1.182 and the Ducks (22 GP) are at 0.7272.

Pts% and Pts/GP are interesting statistics of their own merit (especially looking at averages for teams, divisions, and playoff teams), but what I like about these stats is observing them over time, looking at trends. Sure, the Sharks currently lead Pts% and Pts/GP, but what jumps out to me is that at 4 GP, the Sharks Pts/GP was 0.500(!!), and that the Kings and the Ducks both jumped above 1.50 for brief moments (excluding game one, as should be done).

Since Pts/GP is just Pts%*2, let's take a look at the Sharks' Pts% over 19 games.

Yay! Fun graphics!**

I'm sorry, I couldn't resist.

This graph is basically the exact same as the one above, except the numbers are scaled differently and this one only features the Sharks. Same basic shape of the graph (exactly the same shape of the graph), but without the other teams, it's a little bit easier to see trends for the Sharks. Like, "Wow! Look at that drastic dip! But then luckily the Sharks bounced back to a more normal (for SJ) Pts%."

I like Pts/GP and Pts% graphs better than just simple graphs of Pts vs GP, because on these, losses show up as actual dips rather than just flat portions. But still, there's another graph that could enlighten us even more...

Points%... minus .500

Why minus .500? Basically so I can get the x-axis at .500 (an "average" record***), so I can show the area between the curve and .500. Anything above the axis is above "average," and anything below is pretty terrible, because in hockey, average is not actually .500 (last season only eight teams were below .500 in pts%). The numbers on the y-axis of this graph (and the next couple graphs to come) are pretty irrelevant unless you can quickly add decimals. The important thing to look at is distance from the x-axis. To put things in perspective, last season it took about .620 to make the playoffs. That dip below the axis is pretty ugly for the Sharks, but the fact that they brought their Pts%-.500 back above the axis is promising.

But since .500 isn't going to cut it in today's NHL, shouldn't we look at a graph where the x-axis is more of an "average"?

Yes, so let's start with the Sharks Pts% this season (through 19 GP) with the x-axis at .591 (the minimum Pts% required in the West to make the playoffs last season).

Basically what I did was just move the x-axis up .91. The most noticeable difference is the losing streak. With the adjusted Pts% graph, that losing streak (of just 3 games!) has a much bigger impact on the Sharks maintaining their image as one of the top teams. In other words, it took more wins (compared to the Pts%-.500 graph) to get the Sharks back up to a competitive Pts%. But the Sharks (hopefully!) aren't going to slide into the playoffs as the 8 seed. So how are these Sharks compared to last year's team? Last year the Sharks had a .640 Pts%, so let's place the x-axis there and see how much we can scare ourselves...

Upon first glance... YIKES!

It's actually not that bad. The fact that the Sharks last season were the 2 seed in the West and this graph is currently above the .640 axis says a lot about how well this year's team has done so far. Another thing to look at is the fact that the Sharks, after about game 8, have consistently hovered around last year's pts%, which basically says both teams were/are good.****

Also, the Sharks have only played 19 games so far. It's an 82 game season. Graphs like these will have a lot more meaning in another five or so months. Until then, all of these (and saying the Sharks are either "in a good position" or "in a bad spot") really don't hold much meaning at all.

So with that, see you again for stats later in the year (when they might actually mean something).

*Now that I've pointed out where the Sharks stand, that almost guarantees they'll fall in the standings. Sorry.
**When you open Excel for the first time in years, using a picture as the area under a graph is an exciting thing.
***Average in sports other than hockey, in sports that don't hand out loser points.
****Now can we stop asking the question "Is this year's team better than last?"
We can't answer that until next spring/next summer. Besides, they're both good teams.